Membranes with enhanced biocompatibility will be developed for the immunoisolation of xenogenic islet of Langerhans transplants, and the mechanisms of bioincompatibility will be determined. In a relatively successful technique, a membrane is formed by the coacervation of a polyanionic, water-soluble, nontoxic polymer (algin, a polysaccharide) with a polycationic, water-soluble polymer (polylysine, a polyamino acid). The resulting membrane consists of a hydrogel of mutually entwined, closely interacting negatively and positively charged polymers. Proteins adsorb to this charged membrane and potentiate fibroblast adhesion, spreading, and overgrowth. This process may be initiated or accelerated by inflammation resulting from macrophage spreading and activation on the membrane, which may be stimulated by complement. A novel polymer will be synthesized that has a poly(L-lysine) (abbreviated PLL) backbone with poly(ethylene oxide) (abbreviated PEO) side arms. The polycationic backbone will interact with algin to form a hydrogel membrane, and the nonionic, hydrophilic, PEO side arms will shield the charged surface from the proteins and cells in the biological environment. The permeability and stability of membranes made from coacervates of algin and PLL-graft-PEO will be measured in vitro. Membrane permeance will be assessed by measuring the diffusion of 125I-labeled proteins relative to a mathematical model. Stability in vitro will be assessed by measuring permeance immediately after membrane formation and after 1 month incubation. The biocompatibility of these membranes will be assessed in vitro by measurement of protein adsorption, fibroblast spreading, complement activation, macrophage spreading, and macrophage synthesis of interleukin 1. This will provide insight into the mechanisms of bioincompatibility in these materials. Biocompatibility will be assessed in vivo with islet-free microcapsules in nondiabetic hamsters by measuring the appearance of peritoneal macrophages post-implantation, the attachment of cells to and collagen synthesis upon the microcapsule membranes, and the clumping and incorporation of the microcapsules into the peritoneal tissues. Stability of the capsules in vivo will be assessed by histological examination of peritoneal tissues for broken capsules and by measurement of permeability before implantation and after explantation. Rat islets will be isolated, microencapsulated, and the transplanted intraperitoneally in diabetic golden hamsters. Preliminary in vivo data of 1 month duration with microcapsules made from these materials are presented and lend support to the hypothesis that these microcapsules have enhanced biocompatibility. Microcapsules made with the PLL-graft-PEO material showed dramatically less cell attachment and inflammation than those made with ungrafted PLL.